It's time for the PCPer Mailbag, our weekly show where Ryan and the team answer your questions about the tech industry, the latest and greatest GPUs, the process of running a tech review website, and more!
On today's show, Ryan is back from CES to tackle your questions:
00:30 – What is a core?
02:32 – Where are the FreeSync 2 monitors?
03:35 – Spectre/Meltdown updates for Sandy Bridge?
05:57 – What is Shrout Research?
07:12 – Day-to-day of Shrout Research?
08:13 – Insane GPU prices?
12:14 – Force miners to use Titans?
13:54 – Ryan's fallback career?
15:39 – PCPer website vs. PCPer YouTube channel?
17:58 – Livening up the Mailbag background?
18:43 – Optane SSD as standard storage drive?
20:17 – Hairstyle tips?
Want to have your question answered on a future Mailbag? Leave a comment on this post or in the YouTube comments for the latest video. Check out new Mailbag videos each Friday!
Be sure to subscribe to our YouTube Channel to make sure you never miss our weekly reviews and podcasts, and please consider supporting PC Perspective via Patreon to help us keep videos like our weekly mailbag coming!
Does a MICROCODE update
Does a MICROCODE update require a BIOS flash?
I thought Intel did updates via Microsoft updates too. Didn’t they do that to disable OVERCLOCKING on some CPU’s like the G3258?
Yes and no, it can be patched
Yes and no, it can be patched into a BIOS and loaded early on in the boot process or included in the OS, Microsoft rarely included microcode updates so it’s normally down to motherboard vendors, microcode updates on Linux are readily available.
thumbnail caption:
“omg, only
thumbnail caption:
“omg, only halfway done and I got a turtlehead peeking…!”
Insane GPU prices? not really
Insane GPU prices? not really as that is just what the market demand will do to prices if the supply can not keep up with demand. Now is the right time to get HBM2 on more GPU SKUs and get all that power saved. In fact any power hog memory should be taxed to make it more expensive to use the GDDR# memory than it is to use HBM2 for GPUs/other processors.
Tax GPUs also and use the money to complete more nuke and plants solar/wind/wave power plants for power generation. Nvidia and AMD both need to raise their wholesale GPU pricing so the Retailers will not profit more than the GPU makers themselvs from the GPU damand boom.
All of AMD’s Vega SKUs need HBM2 to act as HBC under the management of Vega’s HBCC and games able to use regular DDR4 systems memory/SSD paged memory as Virtual VRAM. So HBM2 adoption for any Vega graphics(Including APUs that use Vega graphics) can have that HBM2 Cache(HBC).
What is insane is the gamers not paying their fair share for all those expensive GPU R&D investments over the years with Nvidia and AMD using the professional makret’s GPU markups to pay the majorty of the billions needed to fund and maintain GPU development over many years. So Nvidia and AMD need to take the opportunity to raise their wholesale consumer GPU pricing to reflect the actual costs of GPU development. Why shoud the retailers be making bank on AMD’s and Nvidia’s years of hard work and billions in GPU investments.
Consoles may be the ony choice for many who lack the funds to afford GPUs currently and AMD got some market share in the gaming console market and AMD’s Zen/Vega APUs will get more of the integrated GPU market. The more shaders a GPU offers the more value that a GPU will have for compute so it’s not hard to figure that Vega is selling for more than GTX 1080/1080Ti with all those extra Vega nCU based shader cores so valued by the coin miners.
Maybe Nvidia and AMD need to invest more in the DX12/Vulkan Explicit GPU multi-adaptor usage in games/gaming engines so those weaker GPUs that are still available at resonable prices can be scaled up by using more than one of those weaker GPUs for gaming via DX12’s/Vulkan’s Explicit GPU Multi-Adaptor IP built into the new graphics APIs.
That’s one way to get sold those weaker GPUs ganged up for more powerful gaming if Both AMD and Nvidia spend more time getting their lower end GPUs scaling well ubder DX12/Vulkan and explicit GPU multi-adaptor.
Never have I seen so much
Never have I seen so much bollocks written on this sites comment system, the way you write makes it sound like you have a vested interest in GPUs being such a high price.
No interest at all except for
No interest at all except for making the gaming leaches pay their fair share of the GPU’s R&D, development, fabbing and maintainence(driver, firmware, software) costs.
Why should the professional markets be forced to subsidize the gaming leeches so coneumer GPUs should cost more at the wholesale level to give the GPU makers the necessary resources to keep the gaming babies from crying.
GPUs are now as popular for compute as they are for gaming and supply and demand sets their prices at the retail level. The GPU makers can not legally set their products’ retail pricing but the GPU makers need to raise their wholsale pricing so they can fund more R&D and other costs assoicated with GPU creation. Gaming is not the only source of demand for GPUs anymore thanks in part to JHH over at Nvidia and AMD is now able to sell its entire production of GPUs at higher pricing and get some needed revenue to maybe, next GPU generation, offer a line of Gaming only focused GPUs like Nvidia does.
Nvidia’s higher pricing over the years for GPUs has allowed Nvidia to field around 5 different base die tape-outs(GP100, GP102, GP104, GP106, GP107) while AMD has only been able to afford a single base die tape-out(Vega 10) that AMD has used to to create all its professional variants(Radeon Pro WX and Radeon Instinct) and consumer Variants(Vega 56 and Vega 64)
Vega 64 was able to compete with the GP104 based GTX 1080 but not the GP102 based GTX 1080Ti that has 88 ROPs that allow the Ti to really filing out the FPS metrics that keep Nvidia ahead in gaming. Vega 56/64 and the GP104 based GTX 1080 all have only 64 ROPs respectively in which to offer much less pixel fill rates than the GTX 1080Ti.
So AMD needs its GPU SKUs to begin costing even more than Nvidia’s GPUs so AMD can afford to field 5 different base GPU base die tape-outs and like Nvidia has some base die tape-outs that offer up to 96 ROPs so AMD can placate the whining cry baby gamers with that ePeen gamer FPS metrics that the little gaming brats so love.
Now that AMD’s shader heavy GPU designs are popular for coin mining compute AMD should increase its wholesale GPU prices for the long term and get enough revenues to properly compete with Nvidia and Nvidia will always win unless AMD can match Nvidia ROP count to ROP count to match Nvidia in frames flung out there.
It’s great to see AMD’s GPUs commanding higher prices from any market demand and both Nvidia and AMD need to charge more for their products and not let any retailers reap the rewards of years and billions of dollars of AMD and Nvidia investments in GPU technology.
If the little gaming gits can not afford GPUs then let them eat XBONE-X and PS4-Pro! But the ones with the money to pay will be the ones with the GPUs to play, or coin hash.
I’ll bet that AMD sells more Vega 10 dies inside of Radeon Pro WX 9100’s and Radeon Instinct MI25 than AMD sells to the miners and gamers. That project 47 infrencing petaflop supercomputer in a single cabinet takes 80 Vega 10 dies in the form of 80 Radeon Instinct MI25’s and that’s what Vega 10 was really designed for. It also takes 20 Epyc 32 core CPUs/MBs also! So great revenues for AMD’s CPUs and GPUs for the professional market. So boo hoo little gaming whiners that’s supply and demand.
Hey the first Vega 20 base die variant to be fabbed at 7nm is meant for the Super Hot revenue producting compute and AI/Inferencing markets with the Vega 20 die having a ratio of 1/2 DP FP to SP FP. And there you have it from an AMD with such a small gaming market share anyways that AMD is focusing on where the real revenues are, and ditto for Nvidia and the same professional markets. If you have not noticed lately Nvidia’s non gaming GPU/SOC revenues are poised to surpass Nvidia’s Gaming revenues. And JHH hardly talks about GPUs lately at the big tradeshows and events where Nvidia is taliking Compute/AI and automobiles and gaming is more of a not so news worthy event on Nvidia’s mind for the past several years.
GDDR5/5X/6 is a great power hog that needs to be banned for GPUs with HBM2 the only memory allowed for GPU VRAM because of HBM2’s intrinsic power savings. The government needs to Tax power wasting memory to the point where GDDR# memory costs more to use than HBM2 and take the tax money and use that for finishing of those nuke plants and other solar/wind/wave enegry projects. HBM2 will save billions of watthours of enegry and HBM2’s usage needs to be encouraged to help save power and create an HBM2 ecocomy of scale that will see HBM2 costing lower as HBM2 production rises. Tax GPUs and Tax Wasteful(Power hungry GDDR#) memory technology even higher. Make the Gamers share in the cost of their power hungry habits, ditto for coin mining. Let that market fund the construction of thorium plant technology and other energy technology.
So gamers will have to suck it up and work some extra shifts at the pork fat rendering plant! Those gamer freeloading days are over. GPUs will cost whatever the market demand says they will cost and too bad for the gaming little scrubs!
Go suck a porno corks!
Go suck a porno corks! Elitist Phuckstick. You know absolutely nothing of marketing or business. Just another Nerd with an opinion. Nothing more.
If however you do manage to pull your head from your ass ,you’ll understand that fact that gamer leeches as you call them, actually generate cash for both NVidia and AMD. It’s a question of cash flow.
Fork off! Kiss my rosey red gamer be-hind!
AMD does not have that much
AMD does not have that much gaming market share too worry as much about gamers realative to Nvidia and gamers have not/will not pay a proper markup to AMD. But AMD has that GPU compute and that’s getting AMD’s GPUs sold inside of Radeon Pro WX(Vega 10 Die based) and Radeon Instinct(Vega 10 Die based) professional GPU Compute/AI GPU SKUs and the Gamers and Coin Miners will fight(Bid Higher) to get at those Vega 10 dies that do not make the grade for professional use.
AMD has such a small share of the overall discrete GPU market relative to Nvidia and AMD’s Epyc professional CPU sales have more potential to earn AMD more money than Nvidia can earn once AMD gets back to its historical levels of Professional Server/Workststion of CPU market share.
Really AMD could take its GPUs and use that IP for the professional markets and profit more than for any consumer GPU sales, but AMD does need the gameing/coin mining market as a way to get rid of those Vega 10 dies that do not make the WX/Instinct grade to be used in the professional SKUs.
You are just an angry gaming scrub that thinks that both Nvidia’s and AMD’s business models should revolve around consumer and gaming when there is a good business reason for Both Lisa Su(Already Stated that AMD was no longer about low prices) and JHH over at Nvidia(spending all his time talking about AI/Compute and automotive uses for Nvidia’s GPU/SOC, Denver cores, on those auto navigation systems) to focus on the professional compute/AI/Auto products markets with the most revenue growth potential.
The Truth is that AMD does not need to rely on the consumer gaming markets any longer and that’s a good thing from a business standpoint. Nvidia has more to lose from any gaming related affects if the miners take up all of Nvidia’s GPU sales and gamers can not get at Nvidia’s GPUs for gaming. AMD will get more of the integrated graphics markets anyways as those Zen/Raven Ridge APUs get more marketshare and AMD’s discrete GPUs are retailing for more money than even Nvidia’s Flagship GTX 1080Ti, at times, with Vega 64 more valued for its shader cores than any FPS/Gaming usage.
If gamers are not willing to pay the prices that the market sets then gamers are free to eat XBONE-X and PS4-Pro or Nintendo Switch as many have done. AMD and Nvidia are businesses and they will service the markets with the deepest pockets and if they do not then JHH and Lisa can be replaced by actions of their respective shareholders and BODs!
The Nerds will inherit the Earth and Transformers: Robots Got Your Jobs! It’s time for gamers to sell all those GPUs to the miners and make some extra cash to get your gaming fix via the many fine console gaming products on the market until the dust settles! But real Nerds hardly spend their time gaming in the first place.
AMD is not about to be kissing you mangy Red(Neck) Gamer behind and Lisa Su has a PHD so she is at the top of the Nerd world and Raja did his Job well in tuning Vega for the compute/AI markets. And That Vega HBCC/HBC IP will mean that coin miners will not have to ever worry about running out of VRAM should the mining/hashing algorithms need more than 8GB of VRAM because HBM2 can be used as HBC(cache) for up to 512 terabytes of virtual VRAM paged in and out of that HBM2 cache on Vega GPU micro-arch systems that make use of HBM2. AMD is about more than any gaming scrubs with no money and JHH over at Nvidia looks at Lisa with Awe because those Vega GPUs are commanding higher prices than his GTX offerings!
This guy has opinions.
I am
This guy has opinions.
I am not feeling his logic, He must be a part of the problem with gpu prices. Obliviously a European phuckstick to boot. A miner.
How’s about getting a real Job, rather than sucking the mining tit? Oh, but that would require that you’d get your fancypants Nancyboy self into the real world. Not trying to troll you, but I’m tired of your sort.
Not European, but Hey Hey Hey
Not European, but Hey Hey Hey Joe Six-Pack now don’t you have a heart attack! They’ve moved all your jobs on overseas and given them to the Red Chinese. And soon they’ll bring those factories back but they’ll be fully Automated, That’s a Fact!
I’ve worked out of a couple of blue collar uninons in my time and the very thing that destroyed the unions is Joe Six-Packs like yourself there Bubba! Joe Six-Pack and his anti-democratic(not the political party but the idea) ways has more in common with the Ba’athist in Iraq or the fascists. Joe Six-Pack and his Sports as a religion to the detriment of all Knoledge of proper democratic processes and fair play.
The Nerds will fix you real good Joe Six-Pack and you’ll be better off gitting the hell out of the cities and back to the country where your redneck selfs belong!
You stupid Facist Chavs and your Peckerwood ways are not needed any more as the ROBOTS will do that work with no issues 24/7/365. You will pay the proper markup on GPUs or you can just eat gaming console. Your Joe Six_pack days are over now get back behind that plow!
ROBOTS rule and Joe Six-Packs drool! It’s Post Industral for you stinking meatsacks wrapped in moldy cotton rags!
I’m mistaken, He’s a
I’m mistaken, He’s a eco-fascist. Not a miner.
Still a European phuckstick though.
Poor little gaming scrub both
Poor little gaming scrub both AMD and Nvidia have gone beyond the gaming market and the GPUs will be used more for compute/infrencing from now on. That’s GPUs by the thousands for AI infrencing of smartphone/tablet users swipe actions with the phones’ AI chip being trained using thousands of GPUs to be able to tell the user’s mood from finger gestures on those users’ phones and fondle-slabs. Once the user’s mood is established the proper ads to Harry’s Pub will be displayed along with a map and digital coupon for a stiff whisky sour to take the edge of that business professional’s busy day.
There will also be AIs trained to measure FPS gamers’ reflexes and send that information over to the Military Draft Boards so some proper fixes can be applied to those Draft Boards’ random number generators and the little gaming scrubs given some great Reality experience in and around several world-wide Reality Theaters where no respawns can be had should the gaming scrub not get out of the way if the IEDs or RPGs and such soon enough.
Oh how the Nerds will so control things in that Revenge Of The Nerds world of AI/Robotic bliss. The Nerds have even trained the Robots’ AIs to identify the Jocks and the Chavs and to run them down and give them Thermonuclear Wedgies!
@Kareha, that idiots been
@Kareha, that idiots been posting rubbish in the comments section for months, the moment you see a wall of text and the childish name changes it’s best to just ignore it IMO.
Just to piss WhyMe off it’s
Just to piss WhyMe off it’s done. Oh that Daft little WhyMe and his little rusting double-wide down by the abandoned coal mine!
AMD’s GPUs are selling out and those Wholesale GPU prices need to go up up up! Those profesional Compute/AI market rule! Thanks to JHH over at Nvidia for that GPU Professional Compute and Thanks to AMD for the HBM2 that will save on the power used for Compute/AI and coin mining!
Get your a$$ to XBONE-X and PS4-Pro ganming scrubs without the dosh to pay. Nintendo’s Switch also will help there for the scrubs with no jungle in their pockets!
Are all embargoes similar in
Are all embargoes similar in terms of letting the public know when the embargo will be up as far as seeing reviews or is that timeframe also part of the embargo?
I’m looking for new laptop, but wanted to wait for reviews of the Intel Kaby Lake G with Radeon Graphics vs the AMD Ryzen with Vega Graphics for instance. Are you only allowed to provide indirect hints like, you should wait till next month or next week? Or does the embargo actually allow you to specify dates of when we can expect reviews to come out?
Hey you mentioned Pre Mobil
Hey you mentioned Pre Mobil Congress convention, what can we expect from it? 4K at 60fps camera on . That would be nice as hardware will become more faster to process photogrammetry on phone. also is have you guys tested photogrammetry benchmarks on Tensor cores? It seems that one of a good AI problem is photogrammetry.
Thats all thanks.
Hey you mentioned Pre Mobil
Hey you mentioned Pre Mobil Congress convention, what can we expect from it? 4K at 60fps camera on . That would be nice as hardware will become more faster to process photogrammetry on phone. also is have you guys tested photogrammetry benchmarks on Tensor cores? It seems that one of a good AI problem is photogrammetry.
Thats all thanks.
is there a drone rover on
is there a drone rover on mars are they gonna make a geographical map of mars before they put man/woman on mars? like google maps but for mars
thats all
https://www.nasa.gov/press-re
https://www.nasa.gov/press-release/nasa-invests-in-concept-development-for-missions-to-comet-saturn-moon-titan
@Ryan speaking of MWC, Will I
@Ryan speaking of MWC, Will I end up seeing you or Ken over there in a Few weeks?
Also, would love to see a ryzen 5 2400G vs a g4560+ 750Ti video
We’re SOOOOO glad you did not
We’re SOOOOO glad you did not join the USAF!
Thanks for all you do, Ryan & company(s)!!
I apologize for my remarks to
I apologize for my remarks to Market. I usually don’t post rants like that, He rubbed my “rhubarb” the wrong way with his opinion.
That being said, I understand that this is a forum for other opinions.
A free discussion, and a community of Nerds.
Long live the Nerds, and the gamers!
I really want to win that
I really want to win that car!!!!
What’s your thoughts on
What’s your thoughts on cables with circuits built inside, aka thunderbolt, USB type c etc. why can’t they embed all the necessary circuits inside the gadgets, similar to how most other cables are.
When a cable is carrying very
When a cable is carrying very high speed data, the embedded circuit is sometimes needed to tweak the performance of the wires to compensate for the difference between “ideal” wire and what the factory actually used. The motherboard vendor can’t know what’s needed for every type and length of cable that might be connected.
Yes, that’s the explanation
Yes, that’s the explanation I’m hearing everywhere. I’m not saying it’s wrong, but it’s not enough.My issue with that explanation is, those cables are carrying digital signals for a short length. If it’s analog, I can understand. We can still get really high speed data transfer (PCI e type speeds) with just cables ( PCIE extension cables) for up to maybe 4-5 ft. I’m trying to get to the bottom of this because, a device manufacturer ( like Apple) can easily lock out people from using 3rd party (cheaper)cables with those kind of embedded circuitry. So my question is, are they feeding use shit info or is there a legitimate reason. A reason that’s u can’t counter argue ( with reasons 🙂 ).
Will you or any of the PCPer
Will you or any of the PCPer staff be replacing older hardware that will not be receiving firmware updates for Spectre variant 2?
@Ryan Question: did any
@Ryan Question: did any manufacturer ever produce sata-express ssd?
I noticed that motherboard manufacturers stopped putting the ports on the motherboard as of the intel z370 chipset
however they were still present since the z97 chipset till the z270 series chipset -boards.
how come the manufacturers took so long to remove it from production when there never really was’?’ any product to utilize the port to its potential?
also let me know if ya’ll need more cookies. i’ll send some Dutch stroop-waffels next time! =D
In CPUs: what is the point of
In CPUs: what is the point of having a product with a model number and a configurable TDP?* Are all the products with the same hardware essentially the same thing, or is there some binning that goes on?
* Same model but different set TDP means a wildly different performance.